Shadow Education in Urban and Rural China

The social reaction to the new prohibition of private tutoring in the second half of 2021 attracted my attention to the topic of shadow education participation in China. As there have been severe social inequalities between urban and rural China, I look for a better understanding of how much the socio-geographical disparities affected participation in private tutoring before the strict regulations. The research may contribute to future studies investigating urban-rural consequences after implying the prohibition. My research aims to disentangle and interpret the direct and mediating effects of urban-rural disparities on students’ further participation in shadow education services through educational aspirations. An analytical sample of 1783 observations is obtained. Respondents in the sample are students aged 9-22 in mainland China, including 27 provinces (or equivalent). The explanatory variables are retrieved from the individuals, households and family relation datasets in the wave of 2018. The outcome variable “shadow education participation” is retrieved based on the same observations from the individual dataset in the wave of 2020. A binary logit model and a binary probit model for the outcome variable, and two linear probability models for the mediators are constructed. Then the corresponding causal mediation and the sensitivity analysis are performed. The results from the empirical analysis evidenced that students’ socio-geographical school location indeed affects their participation in shadow education. Students’ socio-geographical location as well affects their educational aspirations; however, their aspirations do not substantially affect their participation in shadow education; therefore, it is not a mediator.